Chaos Monkey as a Service and other memorable moments at Velocity Conference

Last week was quite eventful, certainly unpredictable and some would probably say chaotic.

I am a big soccer fan, so with the Copa America and the Euro 2016 tournaments happening, I certainly could scratch that itch. And as always, it is more art than science to predict the results (Iceland winning over England anyone?) and we are treated to memorable outcomes.  Now the world knows there is more to Iceland than being a great place, or even paradise, to build and operate datacenters.

One of the other memorable aspects for me came at last week’s Velocity Conference during the Ignite Sessions when attendees learned that “AWS is Chaos Monkey as a Service”.  As far as predictions go for moving and running applications and workloads in the cloud – it’s happening. See the Cisco Global Cloud Index for some concrete data points and forecasts. Yet with applications based on microservices architectures running on containers, the consumption can be potentially chaotic. This is where DevOps enters the picture helping to bring order to chaos.

Cisco’s Mike Dvorkin gave a keynote talk that I think you will find interesting to watch in case you missed out at last week’s event. Stating that DevOps exists to run applications, but in many cases, the focus is on the micromanagement of every elemental detail – no focus, or understanding, of the application. This breaks down in two fundamental ways. First, there is an explosion in the number of application service definitions.  All of them hardwired for specific use cases, environments, and circumstances related to various stages in the application life cycle. Second is the dependency chaos resulting from a lack of mechanisms to control the consumption of application services. These problems are closely related. The missing piece is how service resources are allocated to their consumers, and how such allocations can be controlled in a consistent, abstracted way – devoid of understanding of instance, architecture, and environmental detail.

Mike discusses the principles and benefits of a controlled consumption model, and its implications to DevOps processes and security in his keynote talk at Velocity Conference. You can watch the recording of his talk here.

Other memorable moments last week included talking to developers, DevOps Engineers and Ops Managers, and listening to how we could help them in their day to day work. We demonstrated that Cisco sponsors several open source projects.  This included a demo of Policy based Infrastructure for Containers with Project Contiv,  Cisco UCS Manager integration with Puppet and a DevOps Model with Cisco and Ansible. Of course, handing out Cisco DevNet T-shirts and having a raffle always proves to be a fairly predictable way to help draw a crowd.

Speaking of crowds, we certainly expect a large one at the upcoming Cisco live! in Las Vegas, July 10-14. Next to visiting the DevNet Zone, I encourage you to plan and schedule your DevOps Sessions.

In case you care about who will win Euro 2016, the final is on July 10.  Since you will be in Vegas, maybe place a bet. Good luck with your prediction on the outcome. I am pretty certain it’s a safe bet Cisco live! will be yet another great opportunity to learn, engage, and make more memorable moments. It may just be organized chaos. See you there!

Image Source: Pixabay

Share